save loop data, matrix- matlab, yalmip

Hi, I'm stuck on generating and keeping the outputs matrix.

As the matrix B contains 't', I should use 'for loop' to generate each solution 'PfeasibleR'

When I run it, only one solution 'PfeasibleR' at t=2 is shown. What I've done was using preallocated variable, PfeasibleR= zeros(2, 18);

Can I get some tips to save value of 'PfeasibleR' per each matrix 'B'?

ops = sdpsettings('solver','mosek','verbose',1,'debug',1);
K= 0.01; %motor torque constant= electromotive force constant
La= 0.5;%electric inductance
b= 0.1;% motor viscous friction constant
J= 0.001;%moment of inertia of the rotor
Va= 140; %armature voltage constant
Ra= 0.5;%armature electirc resistance
PfeasibleR= zeros(2, 18);
 
for t=0.4:0.2:2
A= [-(Ra)/La  0; 0 -b/J];
B= [0; ((K*Va)./(J*Ra))*(1-exp(-(Ra./La)*t))];
D= [0, -K./La; K./J, 0];
I=eye(1);
 
e_1 = 0.8;
y=sdpvar(2,1);
Pr=sdpvar(2,2);
x = sdpvar(2,1);
M=[Pr*A'+ A*Pr+ B*y'+ y*B'+ e_1*D*Pr*D' , y; y' , -e_1];%LMI
 
conR=[Pr>=0, M<=0]; %constraints
obR= -log(det(Pr)); %objective
optimize(conR, obR, ops)
 
PfeasibleR = value(Pr);%(2x2)
yfeasibleR = value(y);%(2x1)
K_R = (yfeasibleR)'/(PfeasibleR);%(1x2)
end
 

Answer 1

[Walter Roberson]

ops = sdpsettings('solver','mosek','verbose',1,'debug',1);
K = 0.01; %motor torque constant= electromotive force constant
La = 0.5;%electric inductance
b = 0.1;% motor viscous friction constant
J = 0.001;%moment of inertia of the rotor
Va = 140; %armature voltage constant
Ra = 0.5;%armature electirc resistance

tvals = 0.4:0.2:2;                    %9
num_t = length(tvals);
all_PfeasibleR = cell(1, num_t);      %1 x 9
all_K_R = cell(1, num_t);
for tidx = 1 : num_t
    t = tvals(tidx);
    A= [-(Ra)/La  0; 0 -b/J];
    B= [0; ((K*Va)./(J*Ra))*(1-exp(-(Ra./La)*t))];
    D= [0, -K./La; K./J, 0];
    I=eye(1);
     
    e_1 = 0.8;
    y=sdpvar(2,1);
    Pr=sdpvar(2,2);
    x = sdpvar(2,1);
    M=[Pr*A'+ A*Pr+ B*y'+ y*B'+ e_1*D*Pr*D' , y; y' , -e_1];%LMI
     
    conR=[Pr>=0, M<=0]; %constraints
    obR= -log(det(Pr)); %objective
    optimize(conR, obR, ops)
     
    PfeasibleR = value(Pr);            %(2x2)
    yfeasibleR = value(y);             %(2x1)
    K_R = (yfeasibleR)'/(PfeasibleR); %(1x2)
    
    all_PFeasibleR{tidx} = PFeasibleR;
    all_K_R_{tidx} = K_R;
end

PFeasibleR_matrix = horzcat(all_PFeasibleR{:}];    %2 x 18

[Junhwi Mun]

Thank you so much!